Research

Public trust in democratic institutions is essential for effective governance, yet global confidence in them has fallen to unprecedented lows. In response, electorates in advanced and emerging democracies are increasingly turning away from career politicians and toward “amateurs”— candidates without prior elected office-holding experience—hoping that these outsiders can disrupt partisan gridlock and elite entrenchment. Using the United States Congress as a critical case, we evaluate these expectations by linking over four decades of election data with 2.2 million records of bill sponsorship and cosponsorship. Contrary to conventional wisdom, we find that electing political amateurs to public office significantly intensifies partisan divisions rather than alleviates them. Amateurs consistently exhibit lower levels of bipartisan reciprocity and perform poorly in attracting opposing-party collaborators to their own bills. This pattern is concerning because cross-party coalitions remain critical for effective legislative governance in the United States and beyond, even amidst well-documented political polarization. We therefore caution against the premise that political outsiders can readily fix political dysfunction, highlighting instead the critical role of experience in sustaining a functional and cooperative democracy.

Do governing elites who engage in undemocratic practices face accountability? We investigate whether American state legislators who publicly acted against the 2020 presidential election outcome sustained meaningful sanctions in response. We theorize that repercussions for undemocratic activities are selective---conspicuous, highly visible efforts to undermine democratic institutions face the strongest ramifications from voters, politicians, and parties. In contrast, less prominent actions elicit weaker responses. Our empirical analyses employ novel data on state legislators' anti-election actions and a weighting method for covariate balance to estimate the magnitude of punishments for undemocratic behavior. The results evidence heterogeneity, with the strongest consequences targeting legislators who appeared at the U.S. Capitol on January 6th, 2021, and weaker penalties for lawmakers who engaged in other antagonism toward democracy. We conclude that focusing sanctions on conspicuous acts against democratic institutions could leave less apparent---but still detrimental---efforts to undermine elections unchecked, ultimately weakening democratic health.

How does majority party security shape reciprocal bipartisan collaboration and influence legislative success? U.S. state legislatures vary widely in the stability of majority control, offering a valuable opportunity for examining how party security conditions the incentives for cross-party collaboration. Insecure majorities may foster reciprocity as both a behavioral norm and a strategic path to legislative advancement, while long-term one-party control can diminish the returns to bipartisan engagement. I develop a theory of selective reciprocity, arguing that majority security fundamentally restructures how legislators engage in and benefit from bipartisan collaboration. Drawing on data from 401,720 bills introduced across 43 state legislatures between 2009 and 2018, I construct novel measures of bipartisan collaboration to evaluate reciprocity. I find that minority party legislators build reputational capital by consistently cosponsoring majority party bills---but their efforts yield few legislative gains in secure majority chambers. Instead, majority legislators selectively reciprocate only on minority party initiatives unlikely to pass, preserving the appearance of cooperation while protecting their policy agenda. By contrast, in insecure chambers, bipartisan cooperation is more likely to produce substantive outcomes. Reciprocity endures but is constrained---selective in form, asymmetric in effect, and structured by the institutional advantages of majority control. These findings raise broader concerns about the marginalization of minority legislators and the limits of representation under conditions of majority security.

Are LGBTQ legislators effective lawmakers? We build on theories that link voter discrimination to legislative effectiveness by arguing that voters’ biases against LGBTQ candidates narrow the candidate pool, leading to the election of only the most experienced and qualified LGBTQ candidates. As a result of this electoral selection effect, we expect that LGBTQ legislators will be more effective lawmakers than their non-LGBTQ counterparts. To test this, we combine data on state legislators’ LGBTQ identification with their State Legislative Effectiveness Scores (SLES). Our findings reveal that LGBTQ legislators are meaningfully more effective than non-LGBTQ legislators. To link our findings to voter discrimination, we leverage over-time variation in discrimination toward LGBTQ individuals. Across four tests, we consistently find that LGBTQ lawmakers elected in high-discrimination environments are more effective than those elected from less discriminatory environments.

Campaigns can feature simple electoral posturing or actual commitments of behaviors that politicians will engage in upon being elected. But can campaigns also offer insights into likely policy outcomes, including those resulting from collective policymaking? To address this question, we take advantage of new scholarship highlighting the enhanced lawmaking effectiveness of bipartisan legislators (Harbridge-Yong et al. 2023). We identify bipartisan campaigners from among more than 800 congressional Representatives. Despite increased polarization, since the year 2000 more than a third of congressional freshmen invoked bipartisan language on the campaign trail. These bipartisan campaigners became effective lawmakers. Their enhanced effectiveness was especially pronounced in Representatives’ earlier terms in office and linked to the lawmaking stages requiring significant coalition-building activities. These findings suggest that campaigns offer voters meaningful insights not only into candidates’ subsequent behaviors regarding the issues they attend to and the legislative votes they take, but also into policy outcomes via their effective lawmaking.

Does bipartisan collaboration enhance legislative success in U.S. state legislatures, as it does in Congress? This article extends Harbridge-Yong, Volden, and Wiseman (2023), who find that members of Congress are more effective lawmakers when they attract a greater share of cosponsors from the opposing party. I adapt their framework to the state level using an original dataset of 401,720 bills introduced across 43 state legislatures between 2009 and 2018. These data enable new, fine-grained measures of bipartisanship, capturing both legislators’ ability to attract out-party cosponsors and their willingness to cosponsor legislation introduced by the opposing party. On the whole, bipartisanship is positively associated with lawmaking success in the states, as it is in Congress. Notably, however, substantial variation across legislatures---such as institutional rules and design, party competition, and majority security--- likely shape the contours of bipartisan collaboration. These findings underscore the value of state legislatures for evaluating how structural features of policymaking environments condition cross-party collaboration and open avenues for comparative institutional research.

The steady professionalization of American state legislatures over the past several decades has created a key tension in political representation: state publics disapprove of professionalized legislatures, on average, yet those legislatures are best equipped to represent their policy preferences. We explain part of this paradox by arguing that citizens’ objections to professionalization stem from distrust of “white-collar” legislators—lawmakers from high socioeconomic classes, who are overrepresented in professionalized chambers. These legislators’ policy priorities are viewed as misaligned with the average citizen’s, which reduces approval more than any opposition to institutional reforms that enhance legislative capacity. A pre-registered conjoint experiment demonstrates support for this claim; citizens do not oppose the institutional expansion of resources for conducting lawmaking. Rather, they react negatively to representation from white-collar lawmakers, whom they associate with professionalized legislatures. Further, we demonstrate in temporal observational analyses of economic outcomes in the states that this opposition is justified. State-level income inequality and poverty have increased in association with the professionalization of state legislatures over time. These findings challenge existing accounts by suggesting that disapproval of legislative professionalism is a rejection of governing by economic elites—not of reforms intended to support legislators and facilitate the process of policymaking.

The Supreme Court’s 2022 decision in Dobbs eliminated the federal constitutional right to abortion. Existing work examines how the decision affected voters’ attitudes and candidates' campaign strategies, but legislators’ reactions remain understudied. We argue that legislators increased their attention to abortion after Dobbs only when gendered representational incentives and party-based electoral incentives aligned. For female Democrats, these incentives reinforce one another, jointly encouraging greater attention to abortion. Female Republicans, however, face gendered representational considerations that encourage attention to abortion, while party-based incentives make such attention electorally costly. Among male legislators, partisan incentives alone are insufficient: although male Democrats have a party-based incentive to increase attention, they lack a gendered incentive. We test this argument by identifying abortion references in nearly 1.6 million statements from U.S. House committee hearings. Difference-in-differences (DiD) estimates show no pre-Dobbs gender or party differences; after the decision, however, female Democrats durably increased their attention to abortion relative to female Republicans, with no change among male legislators.

State legislatures play a critical role in shaping policy across the United States, yet existing measures of legislative effectiveness mask important variation across policy domains. We introduces Issue-Specific State Legislative Effectiveness Scores (ISLES), which leverage transformer-based text classification to assign more than 1.6 million state bills to 18 policy areas. This approach enables systematic, cross-state comparisons of effectiveness at the issue level from 2009 to 2018 across 48 states. Using ISLES, we show how effectiveness varies not only across issues---with some domains advancing smoothly, while others stall in gridlock---but also across legislators, revealing when expertise matters for navigating policy and complexity, how gender influences the advancement of women's issues, and how partisan issue owenership shapes prospects for bipartisan collaboration and policy modernation. By disaggregating lawmaking success by issue area, ISLES provides a novel measure that advances the study of specialization, representation, and polarization in American legislatures.

The efficient production of high-quality, well-developed legislation is crucial to the long-term health of democratic representation. Accordingly, researchers have extensively studied the systematic factors associated with productive, effective lawmaking in the legislatures of advanced democracies. However, this attention has largely focused on the roles of (1) institutional design and (2) politician characteristics, both of which are relatively difficult to manipulate for the purpose of improving legislative output. In this research, we study the efficacy of a direct intervention on productive and effective policymaking and representation: legislator training. We theorize that training lawmakers could enhance the quality of representation via empowerment and/or information. We test these mechanisms by analyzing the impact of training programs in American state legislatures on policymaking and representational outcomes. The first, which aligns with the empowerment mechanism, is Emerge America's efforts to recruit, train, and support Democratic women in 27 states since 2002. Additionally, we test the information mechanism by examining the non-partisan, evidence-based training provided by the Council of State Governments which supplies participating legislators with information and best practices at the start of the legislative session since 1986. The results expand scholars' knowledge of the determinants of effective governance. More generally, this research yields key normative insight into whether democratic institutions can be strengthened through active efforts to enhance legislators' political skills.

Do formal procedural rights granted to minority parties and their members translate into influence in American legislatures? If so, which rules and under what conditions? Political scientists have produced important work on the consequences of legislative rules. Much of this work centers on majority-party advantages – committee gatekeeping powers, special rules, discharge procedures, and motions to recommit or reconsider – while minority-oriented rights are largely understudied. The exceptions are studies of the filibuster, which is generally viewed as a powerful tool of the minority. Beyond that, we lack a systematic account of when and how minority-party procedural rights translate into influence. Drawing on Binder’s (1997) definition of minority rights, we produce a first-of-its-kind dataset of every state legislatures’ minority rights during the 21st century. We then test whether or not legislatures allocating more rights enable minorities to exercise more power. We test for the effect of individual rights on related process outcomes, as well as for the cumulative effect of these rights on legislative outputs. Overall, this paper provides a comprehensive baseline for the scope and consequences of minority-party rights in American state legislatures and informs theories of party power and institutional design.

How do cross-party coalitions form across policy domains, and when does bipartisan collaboration in one area enhance a lawmaker’s broader policymaking capacity? Existing research often treats effectiveness as a single aggregate measure, obscuring variation in how legislators work across different issue areas. I argue that bipartisan collaboration generates transferable reputational benefits: when lawmakers build cross-party credibility in one domain, those reputational gains spill over and improve their capacity to legislate effectively across others. To evaluate this argument, I draw on new issue-specific policy classifications developed in Ballard, Dobson, et al. (2025), which use supervised machine-learning models to classify 1.6 million state bills into detailed substantive categories. I pair these machine-learning–based topic classifications with bill progress and novel measures of bipartisan collaboration to capture lawmakers’ productivity within—and across—distinct policy domains. I then test whether bipartisan collaboration in one issue area predicts heightened effectiveness in others. The results provide clear evidence of reputational spillover: lawmakers who collaborate across party lines in one domain perform significantly better across additional policy areas, even after accounting for partisanship, seniority, institutional position, and specialization. These findings suggest that one way legislators remain productive amidst polarization is by cultivating a bipartisan reputation that expands their capacity to secure policy wins beyond their primary areas of engagement. More generally, the study demonstrates how supervised machine learning can illuminate fine-grained policy dynamics that structure coalition-building in legislative institutions.

When do voters hold lawmakers accountable for their policy choices? Foundational theories of democracy emphasize that citizens can reward or remove elected officials at the ballot box. However, decades of research show that electoral accountability is often limited by information deficits—especially in state legislative elections. We argue that accountability is conditional on issue salience: when a policy is visible, personally relevant, and easily understood, citizens can impose electoral costs on legislators who act out of step with their constituents’ preferences. We test this argument using abortion policy, which has become one of the most salient issues in American politics following the Supreme Court’s Dobbs v. Jackson Women’s Health Organization (2022) decision. Leveraging a unique combination of observational and experimental evidence, we pair district-level results from 2024 abortion ballot measures across ten states with legislators’ abortion roll-call records to evaluate whether—and when—voters sanction misaligned incumbents. Our results reveal clear evidence of accountability: legislators whose abortion voting records aligned with district opinion performed significantly better electorally, while those who diverged faced measurable penalties. We also field a preregistered candidate-choice conjoint experiment in which respondents choose between hypothetical incumbents varying in abortion voting records and other legislative attributes. Across both analyses, the results converge on a central conclusion: electoral accountability in state legislatures is possible—but only when issue salience renders voter preferences clear and politically meaningful. These findings refine classic theories of democratic responsiveness under conditions of limited information in state legislative elections.